How To Distribute Application State with Akka Cluster

Building, testing, containerizing, deploying, and monitoring distributed microservices is difficult, but using Lightbend technologies can make it a lot faster and easier to be successful. In this four part blog series, we walk you through a working Proof of Concept (PoC) built using Lightbend’s open source distributed toolkit, Akka. This PoC delivers a resilient, highly performant, elastic, distributed, and consistent state solution that provides an automatic in memory cache with a persistent backing. Here is the breakdown:

• Part 1 - Getting Started: we walk through building, testing, and running the PoC locally, with instrumentation and monitoring wired in from the very beginning using Lightbend Telemetry.
• Part 2 - Docker and Local Deploy: here we cover containerizing our PoC, and then deploying locally in Docker. Then, we’ll load test and monitor our PoC as we did in the first installment.
• Part 3 - Kubernetes and Monitoring: in this part,we introduce Lightbend Console for Kubernetes (K8s), and then deploy our PoC in Minikube (desktop version of K8s) using YAML files provided in this repository. Again, we’ll load test with Gatling, but this time we’ll monitor our PoC in Lightbend Console.
• Part 4 - Source Code: In our final installment, we do a deep dive into our Scala source code.

---

Update December 6, 2022

• added missing POM dependency for cluster bootstrap on K8s
• added JAVA_TOOL_OPTIONS to Microk8s deployments to accommodate difference when passing environment variables with Maven / Fabric8 to the container.

Update September 19, 2022

• created a new branch (mread-multi-dc-common-journal), which supports Akka's Replicated Event Sourcing "Common Journal" running on Multi-Data Center / Microk8s and Yugabyte. README.md can be found here.

Update September 7, 2022

• update Akka related dependencies to latest
• Learned that sbt uses the env variable JAVA_OPTS to pass configs to Java whereas Maven w/ Fabric8 uses JAVA_TOOL_OPTIONS.

Update September 1, 2022

• Moved to Java 17: converted Actor Protocols and Events to Records
• Akka DNS cluster formation on Docker w/ Yugabyte. Doc here.

Update March 23, 2022

• Complete multi-node testing: integration tests for cluster formation, and gRPC, and http endpoints running under Maven.
• Upgrade Akka, Akka Http, and Akka gRPC to latest.
• multi-node testing defaults to Cassandra, but commenting change be changed to use Yugabyte instead.

Update Feb 15, 2022

• This repository is the Java version of the original Scala version found here.
• Testing Lightbend's Akka Persistence Plugin for R2DBC.

How to Generate Protobuf interfaces and stubs

mvn clean akka-grpc:generate

Running a cluster locally with Cassandra

Start the ElasticSearch Sandbox

Elasticsearch developer sandbox

Start Cassandra in Docker

docker-compose -f docker-compose-cassandra.yml up

Start a cluster node

mvn compile exec:exec -Dapp.configfile="cluster-application.conf"

Start a endpoint node

mvn exec:exec -Dapp.configfile="endpoint-application.conf"

Running a cluster locally with Yugabyte

Start the ElasticSearch Sandbox

Elasticsearch developer sandbox

Start Yugabyte in Docker

docker-compose -f docker-compose-yugabyte.yml up

First time, create the tables

docker exec -i yb-tserver-n1 /home/yugabyte/bin/ysqlsh -h yb-tserver-n1 -t < https://raw.githubusercontent.com/akka/akka-persistence-r2dbc/main/ddl-scripts/create_tables_yugabyte.sql

Start a cluster node

mvn compile exec:exec -Dapp.configfile="cluster-application-yugabyte.conf"

Start a endpoint node

mvn exec:exec -Dapp.configfile="endpoint-application.conf"

Running Multi-Node Tests

1. Start Cassandra first (above)
2. Start the test

mvn test